Keysight Lidar Target Simulator is A Benchtop Solution to Test Lidar Devices in The Design Verification and Manufacturing Stages

From its use in metrology and topography, lidar technology has come a long way to quickly become one of the key technologies in autonomous vehicle (AV) advancements. As automakers progress with real-world testing, it's clear that next-generation sensors offer intriguing features but differ from the silver bullet many thought they'd be at first. Without driver/human inputs, AVs require sensors and computers to work together to read the road and surrounding environment. Most advanced driver aids in the world today use a combination of radar (microwave) and sonar (sound) to warn about unseen threats and help stop a vehicle before a collision occurs. Lidar can perform similar functions, developing an image of the environment around it while still representing the best option for AVs to see the light.

Lidar Technology in Automotive

The automotive industry is undergoing an existential transformation with autonomous driving (AV) and advanced driver assistance systems (ADAS). To achieve a fully autonomous driving experience, sensor fusion is positioning itself as the key enabler in the automotive industry. While sensor fusion focuses on software algorithms combining the data collected from cameras, radar, and lidar to obtain meaningful decision-making information, data quality is paramount. It’s true that there is still a long way to go in mimicking the human senses and brain for level 5 driving automation, but sensor fusion is a further step in that direction.

Cameras and automotive radar have been the pillars for up to level 2 driving automation. Level 3 and higher require additional sensors and redundancy. For example, car manufacturers use radar in various systems – blind-spot monitoring systems to detect vehicles before lane change, automatic emergency braking systems to stop a vehicle before it touches an obstacle, and adaptive cruise control to maintain a consistent distance between two cars. Automotive lidar brings up high resolution and accuracy, low light effectiveness, and 3D mapping to the other two sensors. Lidar tracks obstacles and vehicles to maintain safe distances; it helps identify road signs, traffic signals, and road markings for real-time hazard analysis, ensuring autonomous vehicles’ effective operation. However, automotive lidar is yet to go mainstream due to its cost.

Lidar 101

Lidar, short for "light detection and ranging", is a device that uses laser pulses for detection and ranging; it can operate at different wavelengths like 850nm, 905nm, 940nm, and 1550nm. The lidar sensor comprises a laser source acting as a transmitter, a photodetector acting as a receiver, and an assembly of lenses or optics to steer and collect the laser pulses. When the pulse touches an object, it bounces back to the lidar unit. The system then receives the pulse and calculates the distance with the object based on the time elapsed between the emission of the pulse and the reception of the return beam. As the beams return to the system, they begin forming a picture of the vehicle's surroundings and use computer algorithms to piece together shapes for cars, people, and other obstacles. Lidar's use of pulsed lasers allows it to map the 3D model of an environment quickly and more accurately than radar or sonar.

There are two types of lidar:

• Time of Flight (ToF) lidar: uses the time difference between transmitted and reflected laser pulse to calculate the target's distance, also called range.

• Frequency Modulated Continuous Wave (FMCW) lidar: uses frequency difference between transmitted and received modulated laser chirp to calculate distance and derive the target's velocity. FMCW lidar uses coherent detection that helps achieve improved range resolution and the ability to measure dim and bright targets simultaneously.

We can also categorize lidars based on detection or scanning technology.

Lidar Test Challenges

As with radar-based systems, lidar sensor makers must ensure their systems quickly and reliably detect objects, enabling advanced driver assistance systems (ADAS) to work correctly before being commercially deployed. To properly test sensors, designers must often depend upon large floor spaces and traditional target boards for range and reflectivity tests. The industry also faces challenges in reducing sensor costs and scaling to mass production.

The main challenge for the mass adoption of lidar by automotive OEMs is bringing down the cost of lidar sensors and possibly making them comparable to radar. The price of lidar devices has decreased significantly in the last few years; however, there's a scope to make it more competitive. Many factors are driving the cost up, such as R&D, material and production costs, and lower production volumes.

KEYSIGHT lidar target simulator (LTS)

Keysight lidar target simulator is a benchtop solution to test lidar devices in the design verification and manufacturing stages. The solution is designed to reduce the test cost by significantly reducing the floor space as it can simulate target distances from 3 to 300 meters and surface reflectivity from 10% to 94% while supporting volume production and accelerating the testing with full automation and analytics software.

The solution comprises two main building blocks and an optional third-party cobot: an LTS base unit, an LTS remote optical head, and a cobot that is optional with a mounting platform.

Figure 1. Keysight E8717A Lidar Target Simulator

The LTS base unit is the heart of the test system, where distance and reflectivity simulations are performed. The remote optical head enables the base unit to receive the signal from the lidar sensor and transmit it back delayed (distance simulation) and attenuated (reflectivity simulation). In addition to target distance and reflectivity simulations, the Keysight LTS solution is equipped with advanced features that enable automotive lidar makers to gain additional insights into the sensor performance. The solution allows for lidar testing under different conditions with physical target boards.

In addition to these benefits, Keysight LTS solution has an automated setup calibration algorithm that enables users to perform accurate and repeatable tests for lidar sensors. Opto-mechanical alignment is performed to align the centers of the lidar sensor and remote optical head. This setup calibration and alignment methods ensure accurate distance and reflectivity simulation.

A lidar sensor is tested for all its specifications and performance parameters in the design verification stage, which includes lidar calibration and lidar verification. The requirements for calibration are defined depending on the lidar sensor's optical design, which the Keysight LTS target simulation capability and mechanical and optical alignment and setup calibration performance can accomplish. Then, for verification, the automotive lidar sensor testing occurs for range (minimum and maximum), range accuracy, azimuth and elevation angle accuracy, and field of view tests.

In the manufacturing stage, end-of-line testing ensures the lidar sensor is functional before shipment. These end-of-line manufacturing tests are typically a subset of verification tests. For an efficient yield and high-volume production, the solution shortens test time with software automation enabled by Keysight PathWave Test Executive for Manufacturing (PTEM).

Going forward

Automotive lidar is a promising step toward a fully autonomous driving experience. With sensor fusion and more precise algorithms, there will be higher redundancy of meaningful data streams. Data that one day will help reconstruct a decision-making process close to the level of complexity that a human brain can handle.